Impacts of De‐NOx system layouts of a diesel passenger car on exhaust emission factors and monetary penalty

Automobile emissions are significantly dependent on the after‐treatment system performance, which is partly determined by exhaust temperature. Regarding diesel passenger cars, after‐treatment systems generally include diesel oxidation catalyst (DOC), diesel particulate filter (DPF), and selective catalytic reduction (SCR). The layouts affect their temperature variations because of the heat loss and thermal capacity of tailpipes and after‐treatment systems. As for the original layout of DOC+DPF+SCR, nitrogen oxides (NOx) emissions are the main concerns of diesel vehicle emissions, especially under cold‐start conditions. Ammonia Creation and Conversion Technology (ACCT) system shows excellent performance of reducing cold‐start NOx emissions; additionally, the damage costs of individual exhaust emissions are different greatly, which may change the priority of emission reductions when considering monetary penalty. In this article, the impacts of the after‐treatment system layouts on the exhaust emission reductions were investigated based on a diesel passenger car; additionally, SCR and ACCT systems as the De‐NOx devices were adopted individually in corresponding scenarios; the after‐treatment system layouts were assessed from the viewpoints of both emission factors and monetary penalty. The results indicated that the ACCT system presented much better NOx reduction effectiveness than SCR system over different layouts. NOx reduction efficiency was very sensitive to vehicle operation conditions over the upstream layouts of NOx reduction devices. The layout‐1 of DOC+ACCT+DPF showed the lowest global emission factors from the diesel passenger car. DPF was much easier to achieve regeneration under the original layout conditions due to its shortest distance to the engine. The layout‐2 of ACCT+DOC+DPF had the minimum monetary penalty factor of exhaust emissions from this diesel passenger car.


| INTRODUCTION
With the increase of the amounts of automobiles, the problems caused by exhaust emissions are becoming more and more prominent. [1][2][3] As indicated by the report, 4 carbon dioxide (CO 2 ) emission from transportations account for 29% of the total CO 2 emission, being the largest proportion, followed by electricity, industry. Nitrogen oxides (NO x ), particulate matter (PM 2.5 ), and volatile organic compounds (VOCs) from automobiles contribute to large proportions of total hazardous emissions, being approximately 35%, 5%, and 20%, respectively, indicated by Minnesota Pollution Control Agency. 5 Emission regulations aiming to drop automobile emissions are becoming more and more stringent 6 ; additionally, real driving emissions (RDE) test procedure is also introduced. 7 Compared with standard driving cycles such as Worldwide harmonized Light vehicles Test Cycles (WLTC), 8 the real driving conditions are more complicated and variable, which leads to higher emission factors over real driving than WLTC. 9,10 The issue of higher RDE than laboratory emissions is mainly for CO 2 of all vehicles and NO x of diesel vehicles. It was demonstrated that Euro 6b diesel cars emitted emit 5-16 times of NO x in the real world than certification tests 11 ; meantime, CO 2 emission was 1.4-1.8 times of the certificated values. CO 2 emission of a Euro 6 gasoline passenger car was approximately 194 g/km under realworld driving conditions, being much higher than that of standard driving cycles (lower than 100 g/km). 12 Previous studies addressed the importance of reducing real-world CO 2 and NO x emissions. Yang et al 13 also demonstrated that NO x and CO 2 emissions were critical to address RDE of Euro 6 passenger cars.
With regard to the current techniques mitigating automobile emissions, the effective measures aim to increase after-treatment efficiency [14][15][16] and decrease emission formations by controlling in-cylinder combustion. [17][18][19] As for the current mature technologies, in-cylinder control usually brings about fuel penalty, although it can drop emissions to a low level. In order to meet stricter emission regulations, after-treatment systems are indispensable to balance the emission formations and fuel consumptions. One of the most important factors correlating the aftertreatment system efficiency is temperature. 20 The effectiveness of catalyst-based after-treatment systems is excellent under high exhaust temperature conditions where the emission reduction efficiency can be higher than 90%. 21,22 However, their performance under cold-start conditions is poor. 23 According to Gao et al's study, 24 exhaust emissions during engine warm-up process accounted more than 50% of the total emissions over the given trips. Heeb et al 25 demonstrated that approximately 70% of the total carbon monoxide (CO) emission over Federal Test Procedure (FTP) (1400 s in total) was caused by the first 500 s of the trip. Research 26 compared exhaust emission rates of different vehicles, demonstrating that the emission rates over cold start were much higher than hot start, with the maximum difference being higher than 100 times of hot-start emissions. Additionally, the emissions were significantly worsened by low ambient temperature for passenger cars meeting different emission regulations due to much heat loss by tailpipe and upstream after-treatment systems. 27 It was consistent with the conclusions of the reference 28 which addressed unregulated emissions as well. Du et al 29 researched the RDE under cold-start conditions, revealing that the cold-start period accounted for a significant proportion of the total urban driving emissions; meantime, the cold-start emissions were very sensitive to driving behaviors; additionally, the correlations between coldstart emissions and ambient temperature were addressed. The impacts of cold-start emissions by driving behaviors were also discussed in Gao et al's study. 30 Many technologies are investigated to drop cold-start emissions such as burners, 31 reformers, 32 thermal insulations, 33,34 heat storage materials, 35 electric catalytic converters (eHC), 36 Ammonia Creation and Conversion Technology (ACCT), 37 Ammonia Storage and Delivery systems (ASDS) 38 ; meantime, some of these measures are helpful for diesel particulate filter (DPF) regeneration. The main purposes of these technologies are to increase exhaust temperature or provide an easier way to supply reactants (eg, gaseous ammonia), ensuring high efficiency of after-treatment systems. Gao et al 30 discussed their effectiveness, energy penalty, and complexity. Regarding the common layouts of after-treatment systems (DOC+DPF+SCR), NO x emissions are the main concerns. ACCT system as a new technique aiming to drop cold-start NO x emissions has excellent performance and no energy penalty. 37 It used exhaust energy to hydrolyze AdBlue into gaseous CO 2 , water (H 2 O), and ammonia (NH 3 ) under high exhaust temperature conditions; then, they were formed as a new product (ACCT fluid) after cooling down. ACCT fluid was much easier than AdBlue to provide gaseous NH 3 under relative low-temperature conditions.
One of the vital factors affecting SCR efficiency was the exhaust temperature, 39,40 which is related to the engine operation conditions and the distance of the SCR system to the engine. For the after-treatment systems being the downstream of others, there was a temperature drop and lag caused by heat loss and thermal capacity of the tailpipes and upstream after-treatment systems. 41,42 Lao et al 43 investigated the impacts of after-treatment system layouts on diesel engine emissions. It indicated that DOC+DPF+SCR layout was found to be more beneficial than DOC+SCR+DPF under specific conditions. DPF-infront system was more robust regarding the changes in emission regulations. Gurupatham et al 44 also researched the effect of after-treatment layouts on the global performance of DOC, DPF, and SCR. Moving SCR to the downstream of DPF led to the drop of NO x reduction under cold-start conditions; however, it was benefited for DPF regeneration. He et al 45 simulated the after-treatment system performance over various layouts, with the results indicating that the individual after-treatment systems had correlative impacts on their emission reduction efficiency.
In fact, the reductions of some individual emissions from vehicles should have priorities determined by social, human health, and economic impacts, which are considered in the damage costs. The damage cost of NO x emissions was much higher than CO and hydrocarbon (HC) 46 ; additionally, SCR system was placed the downstream of DOC and DPF systems for the common layouts. It caused much NO x emissions from cold start and warm-up process. When considering the emission reductions from the perspective of monetary penalty, SCR system should have a higher priority than DOC in after-treatment layouts. DPF efficiency depended less on the exhaust temperature, but the regeneration has the requirements of high temperature. Taken the monetary penalty into consideration, the optimal layouts of the after-treatment systems would be different.
According to the government report, 46 the damage cost of the individual air pollutants is different. Regarding the regular exhaust emissions from internal combustion engine vehicles, particulate matter (PM) has the highest damage cost, followed by NO x and HC 46 ; the damage cost of PM is thousands times higher than HC; additionally, the damage cost of CO is much lower than others. 47 The applications of monetary penalty factors calculated from emission factors and corresponding damage cost can be used to assess the impacts from vehicle exhausts. To the authors' knowledge, this is the first time to estimate the monetary penalty factors of exhaust emissions from passenger cars over various after-treatment layouts. In this study, emission characteristics of a diesel passenger car were investigated using a numerical simulation method when the diesel passenger car was equipped with different De-NO x devices; the effectiveness of the after-treatment systems were discussed under various layout conditions; meantime, the monetary penalty factors of the exhaust emissions were used to assess their overall impacts on environment, human health, and society, which provides a new insight to regulate the exhaust emissions.

| MATERIALS AND METHODS
In this section, the simulation models were described in detail, the vehicle model and ACCT model were validated using experimental data, and various layouts of aftertreatment systems were provided.

| Simulation model description
A vehicle simulation model was set up using GT-SUITE software, including a powertrain model, an after-treatment system submodel, and an emission submodel.

| Descriptions of the vehicle model
The vehicle used in this investigation was a Euro 6 compliant diesel passenger car whose power source was a four-cylinder, four-stroke, turbocharged diesel engine. The specifications of the diesel vehicle are shown in Table 1. The maximum power output of the diesel vehicle was 103 kW, corresponding to the engine speed of 4000 rpm. The compression ratio of the diesel engine was 16.5. The simulation model of the diesel passenger car is shown in Figure 1. This vehicle model included three modules, which were vehicle powertrains, emission sources, and after-treatment systems. Engine model, transmission model, and control model were included in the powertrain system; the engine model was based on experimental tests, including brake-specific fuel consumption and brake mean effective pressure maps. In order to consider the effect of cold start on pipe-out emissions, the maps of exhaust temperature, emission factors, and exhaust flow rates were included in the emission model. With regard to the after-treatment system, it included a DOC, a DPF, and a SCR. The DOC was followed by a DPF, being at the upstream of a SCR device.

| Descriptions of aftertreatment systems
The main concerns on the gaseous exhaust emissions from diesel passenger cars were NO x . In order to meet stricter emission regulations, for example, Euro 7, the SCR system could be replaced by an ACCT system to further decrease NO x emissions by effectively delivering gaseous ammonia under low exhaust temperature conditions. The operation theory of the ACCT system was reported in the reference. 37 The ACCT system was consist of an AdBlue tank, an AdBlue pump, an ACCT reactor, an ACCT fluid tank, and an ACCT fluid injector (see Figure 2). AdBlue was injected into the ACCT rector under high exhaust temperature conditions; then, AdBlue was hydrolyzed into CO 2 , gaseous water, and gaseous ammonia under the assistance of exhaust energy; furthermore, the gaseous hydrolyzed products were cooled down in the ACCT tank, being converted into ACCT fluid (a new product). The hydrolysis temperature of ACCT fluid was much lower than AdBlue such that ACCT fluid was more effective to drop NO x emissions under low exhaust temperature conditions, such as urban driving, stop-and-go scenarios.

| Descriptions of the after-treatment system layouts
Layouts of the after-treatment systems affected the emission reduction efficiency, DPF regeneration, and the energy penalty. The scenarios of the after-treatment system layouts in this investigation are shown in Figure 3. Due to high effectiveness of ACCT system dropping NO x emissions, it may generate impacts on the layouts of aftertreatment systems in terms of optimal global performance. In this investigation, three different after-treatment layouts were provided. The changes of after-treatment layouts inevitably affected after-treatment efficiency due to the temperature changes accordingly; additionally, it would influence DPF regeneration processes, which required high exhaust temperature. Regarding the DPF regeneration, passive regeneration using nitrogen dioxide (NO 2 ) or catalyst was adopted in the work; and the difference

| Validations of simulation models
In order to estimate the vehicle emissions with high precisions, the vehicle model including the fuel consumption and exhaust emissions over original after-treatment layout was validated using experimental tests.

| Validations of vehicle model
Validations of the vehicle model using fuel consumption were based on WLTC, as shown in Figure 4. Simulation results over high vehicle speed matched better with the test results than low vehicle speed conditions. Over low vehicle speed conditions, the fuel consumption rates were low such that small errors would cause a significant difference between simulation and test results. Only small amounts of simulated results showed notable deviations with the test results. Figure 5 compares NO x emission rates of experimental and simulation results for both engine-out and pipe-out emissions. The difference between experimental and simulation results was minor, which indicated high precisions of the emission models and SCR model (including the chemical reactions). Concluded from Figures 4 and 5, the vehicle simulation model was accurate to do the future research. As shown in Figure 5, NO x formation was the highest over highspeed and extra-high-speed regimes. NO x formation rates under low-speed conditions were quite low due to low in-cylinder combustion temperature. However, the pipeout emission rates were almost zero over high-speed and extra-high-speed regimes, benefiting from high NO x reduction efficiency. At the start of the driving cycle, the exhaust temperature was low, leading to low SCR efficiency. The main focus of the current enhanced NO x reduction techniques was cold-start regimes.

| Validations of ACCT system
ACCT system was validated using experimental tests separately. The boundary conditions of the ACCT simulation model were the same as the ones in the work described previously. 37 The boundary conditions of the simulation and test were: hot air flow rates, 250 kg/h; ACCT fluid flow rates, 400 g/h; air temperature at 100°C, 150°C, 200°C, 300°C, and 400°C; ammonia (NH 3 ) monitoring point, 800 mm downstream of the ACCT fluid injector. Each scenario lasted 30 s in the experiment. 37 Comparisons of simulation and test results are shown in Figure 6. As indicated in the work of Wilson et al, 37 ACCT fluid was almost fully hydrolyzed over 100°C; however, the tested NH 3 concentration was slightly lower than the simulation results, which was caused by NH 3 films forming on the walls of the pipes used in the test. If the test lasted longer, NH 3 concentration would approach 500 ppm after the NH 3 absorption by the pipe walls was saturated. The precision of the ACCT simulation model was acceptable for the future investigation.

| RESULTS AND DISCUSSION
In this section, exhaust emissions and monetary penalty factors over different after-treatment system layouts and DPF regeneration strategies were explored; additionally, the De-NO x measures including the normally used SCR technique and the ACCT system were adopted in

| Exhaust emissions over different after-treatment system layouts
In this section, exhaust emissions including CO, HC, NO x , and PM over various after-treatment layouts were reported when the diesel passenger car was equipped with different De-NO x systems, namely SCR and ACCT systems.

| Exhaust emissions over various layouts
Efficiency of the DOC, SCR, ACCT, and DPF regeneration were significantly dependent on the exhaust temperature, which would be affected by vehicle operation conditions, thermal capacity, and heat loss by tailpipes and upstream after-treatment systems. The exhaust temperature at different positions of the pipeline is shown in Figure 7. The temperature presented in this figure was corresponding to the original layout of after-treatment systems. As can be seen, the exhaust temperature before DOC changed significantly with vehicle operation conditions, and it had a similar pattern with vehicle speed. High vehicle speed usually led to high exhaust temperature, although the temperature was more relied on vehicle accelerations. At the start of the cycle, the after-treatment temperature was low, leading to the formation of N 2 O emission in SCR. According to Yang et al's study, 48 the exhaust temperature before after-treatment systems presented high dependency on the vehicle speed under real driving conditions. The temperature before DPF was much smoother than that before DOC due to heat loss and thermal capacity of upstream systems. Because of the large volume of DOC and its thermal capacity, there was a temperature lag before DPF. Thermal capacity of the DOC device could alleviate Vehicle speed/ km/h Speed the temperature changes caused by the heating and cooling effect. The temperature profile before SCR device had a similar pattern with that before DPF, but it had an approximately 100-s delay. According to the reference, 20 the exhaust temperature of SCR system was much smoother than the patterns of vehicle speed, which agreed with the authors' results. It was also supported by the results 49 under real driving conditions where smooth temperature profiles were observed after after-treatment systems. As for this diesel passenger car over WLTC, it took more than 200 s for the temperature before SCR to reach the light-off value. Regarding a heavy duty diesel vehicle under real driving conditions, the warm-up durations lasted more than 1000 s where the temperature before SCR was lower than 100°C. 50 There were lots of vehicle stop events where the temperature dropped significantly due to the idling. In the real-world tests, the temperature distributions changed significantly with vehicle driving situations. 50 Over most of the operation time, the engine was under low engine load conditions. Distributions of NO x emission rates over different aftertreatment system layouts for SCR scenarios are shown in Figure 8. Proportions of high NO x emission rates dropped significantly regardless of the after-treatment system layouts. Layout-2 had the highest proportions of low NO x emission rates, followed by layout-1 due to their shorter distance to the engine than the original layout. The proportions of NO x emission rates being higher than 0.01 g/s were nearly zero for layout-1 and layout-2. Distributions of CO and HC emissions over various scenarios when the vehicle was equipped with SCR/ACCT are presented in Figures S1 and S2. Since SCR and ACCT systems only affected NO x emissions, CO distributions over SCR and ACCT scenarios were almost the same. The proportion of engine-out CO emission rates being smaller than 0.004 g/s was quite low; and the proportion decreased generally with the increase of emission rates when CO emission rates were higher than 0.004 g/s. Regarding pipe-out CO emission rates, the distributions over the original layout were the same as layout-1 because the DOC position was the same. The CO emission rates were mainly focused on the values being lower than 0.002 g/s. Compared with original layout and layout-1, the proportion of high CO emission rates increased slightly for layout-2. Under the impacts of after-treatment system layouts, the HC distribution patterns were similar to CO.
For different after-treatment system layouts, the difference in pipe-out NO x emission rates was mainly caused by SCR efficiency, which is shown in Figure 9. It was obvious that SCR efficiency was low at the start of the driving cycle, especially for original layout and layout-1. SCR efficiency of layout-1 had a similar pattern with original layout, which was consistent with the temperature evolutions ( Figure 7). Because SCR system was the closest to the engine for layout-2, the temperature was higher than the other two layouts at the first 150 s, resulting in higher SCR efficiency. SCR efficiency was almost 100% for the three different layouts after 900 s. The efficiency was dropped significantly for layout-1 around 600 s and original layout around 700 s due to the temperature decrease in those periods, which was caused by low vehicle speed around 550 s. Figure S3 shows the proportions of CO reductions by DOC over various after-treatment system layouts. CO reductions reached nearly 100% in a short time after vehicle started for original layout and layout-1; however, it took more than 200 s for layout-2 to achieve high CO reduction efficiency. HC presented a similar pattern (see Figure S4), but the fluctuations of HC reduction efficiency were higher than CO. According to Boriboonsomsin et al, 50 the time-average NO x reductions by SCR were lower than 80% for majorities of the vehicle groups. Exhaust emissions were also sensitive to ambient temperature, especially for NO x emissions which were increased from approximately 0.09 g/km to 0.67 g/km over WLTC when the ambient temperature was decreased from 23°C to −5°C. 51 Lower ambient temperature meant more heat loss. The temperatures used in Ko et al's study 51 were in the common ranges for many countries. It implied that the emissions would be worse in winter. Due to lower light-off temperature of ACCT system than SCR system, NO x emission rates over ACCT scenarios were lower than corresponding SCR scenarios. Proportions of NO x emission rates being higher than 0.0075 g/s were quite low, as shown in Figure 10; NO x emission rates were mainly focused on the regions being lower than 0.005 g/s. Figure 11 shows the ACCT efficiency of NO x reductions over various after-treatment system layouts. Compared with SCR scenarios, the lowefficiency regions were much less for ACCT scenarios. The "zero-flat" durations for the original layout and layout-1 were shorter than corresponding SCR scenarios. Regarding layout-1, the efficiency drop around 680 s was weakened by comparing with SCR scenarios, and the efficiency drop around 750 s was limited for original layout. Such excellent performance of ACCT system was also demonstrated by Gao et al's study. 30 The comparisons of the average ACCT and SCR efficiency over different aftertreatment layout are shown in Table 2. The efficiency improvement of the SCR and ACCT efficiency by the layout were significant; meantime, the efficiency improvement by enhanced De-NO x system was the highest over the original layout among the three scenarios.

| DPF regeneration over different after-treatment layouts
Different from DOC and SCR/ACCT systems whose efficiency was significantly dependent on the exhaust temperature, DPF efficiency was mainly relied on the designs and the engine operation conditions (PM concentration and exhaust flow rates). DPF regeneration process was dependent on the exhaust temperature. Figure 12 shows DPF efficiency over NO x assistant regeneration scenarios. Under relative low-temperature conditions, chemical reactions between NO x and DPF could occur, 52 which achieve the continuous regeneration of DPF. DPF efficiency was almost the same over different layouts in the first 1100 s; the difference between the original layout and layout-1/layout2 was slightly enlarged in 1100 s-1800 s. It was mainly caused by the DPF regeneration process. As indicated by Meng et al, 53,54 the particle number and average particle diameter increased during the DPF regeneration process. It was also consistent with Bensaid et al's opinion. 55 Under the NO x assistant regeneration conditions, it could avoid much PM depositions. The difference of DPF efficiency between NO x assistant regeneration and catalyst assistant regeneration was limited over individual after-treatment layouts (see Figures S5 and S6). Regeneration process of DPF system could be reflected by the amount of soot depositions on filters, as shown in Figures 13 and 14. As can be seen, the amounts of soot depositions increased continuously if without any regeneration measures, which could led to the significant increase of the engine backpressure. 56 The regeneration process started from around 300 s for the original layouts; however, it was around 1000 s for the other two layouts. When the exhaust temperature was higher than 200°C, the soot oxidation rates increased significantly. 52 By analyzing the test data from Jiao et al and Rossomando et al, 52,57 DPF regeneration temperature over catalyst scenarios was higher than NO x scenarios; however, the catalyst assistant regeneration was slightly easier than NO x assistant regeneration for all the after-treatment layouts in this study. Due to the existing of NO x in exhaust, the catalyst assistant regeneration also included the impacts from NO x assistant regeneration in theory. Regarding the original layout, the temperature before DPF was higher than the other two layouts (see Figure 15) because the DPF was closer to the engine. As for the DPF regeneration, DOC+DPF closed couple strategy was more efficient.

| Exhaust emission factors and monetary penalty factors
Exhaust emission factors over WLTC were investigated under various after-treatment system layouts, and results are shown in Tables 3 and 4. As can be seen from the tables, the exhaust emission factors were lower than the limitations of Euro 6 emission regulation for both SCR and ACCT scenarios. Without considering the DPF regeneration problems such as energy needed, layout-1 and layout-2 presented better performance than the original layout regarding emission factors, which was caused by the improved overall thermal status of after-treatment systems. Layout-2 showed the lowest NO x emission factor, although the CO emission factor was doubled. It should be noted that PM emission factors over NO x assistant regeneration scenarios were slightly lower than catalyst assistant regeneration. Compared with layout-1 and layout-2, the original layout presented the highest PM emission factors due to insufficient NO x emissions (in DPF) which conduced to oxidizing PM. Due to lower hydrolysis temperature of ACCT fluid, NO x emissions factors over ACCT system were much lower than SCR scenarios. SCR system and ACCT system presented limited impacts on PM emission factors. It was indicated in the reference, 20 SCR efficiency reached the maximum value under the exhaust temperature of 350°C, and the efficiency dropped gradually with increasing temperature. Additionally, the optimal exhaust temperature for SCR efficiency changed with the types of catalyst coating on SCR. 50 Regarding the upstream layouts for SCR system, the exhaust temperature was lower than 400°C over WLTC; however, exhaust temperature could be much higher when the vehicle operated at high load conditions, such as operating under the uphill conditions for a long period in real driving conditions. SCR system should be avoided the closest distance to the engine. Under the real driving test, average NO x emission factors for Euro 6 compliant diesel vehicles was approximately 0.31 g·km −1 in urban areas; the value was around 0.2 g·km −1 for motorway. 58 The RDE was much higher than WLTC results; additionally, the emission factors over urban driving were higher than motorway. It addressed the importance of high efficient after-treatment systems and the challenge of lowering NO x emissions under low-speed conditions. As for the real-world driving, 59 low vehicle speed conditions had the highest emission factors, especially for NO x emissions. Based on the authors' previous investigation, 14 eHC technique was used to decrease the exhaust emissions during cold-start conditions. The effectiveness of eHC was higher than ACCT system; however, eHC brought about much energy penalty which was inconsistent with the low-carbon driving requirements.
Due to the differences in damage cost of the individual exhaust pollutants, the monetary penalty factors of exhaust emissions would be much different from exhaust emission factors, as shown in Tables 5 and 6. The damage cost of individual exhaust emissions from government report is shown in Table S1. Due to high damage cost of NO x and PM emissions, the monetary penalty factors of NO x and PM emissions were much higher than CO and HC emissions for both SCR and ACCT scenarios. Original layout presented the highest monetary penalty factor, and layout-2 had the lowest factor, regarding both individual and total monetary penalty factors. The factors under ACCT scenarios were much lower than SCR scenarios. DPF regeneration approach showed minor impacts on total monetary penalty factors. NO x emission should have the priority in order to drop the monetary penalty factors, followed by PM because of their high proportions in total factors.

| CONCLUSIONS
In order to investigate the impacts of after-treatment system layouts on exhaust emissions and corresponding monetary penalty factors, a diesel passenger car was simulated over WLTC; additionally, the effect of De-NO x systems (SCR and ACCT systems) on the exhaust emissions and monetary penalty factors were discussed. The main conclusions are as follows: T A B L E 4 Exhaust emission factors when the vehicle was equipped with an ACCT system 1. After-treatment temperature was sensitive to the vehicle speed when it was close to the engine; the longer the distance of the after-treatment systems to the engine, the smoother and lower the temperature profile was. For SCR and ACCT scenarios, NO x emission distributions presented similar patterns over various layouts; however, ACCT scenarios showed a lower proportion over high emission rate regions. 2. ACCT scenarios presented higher NO x reduction efficiency than SCR scenarios under different layout conditions, especially the first 800 s of WLTC NO x reduction efficiency over original layout and layout-1 showed similar patterns, but there was a lag of approximately 130 s for the original layout. As for the DPF regeneration, original layout was much easier than layout-1 and layout-2, caused by higher DPF temperature. 3. ACCT system showed lower NO x emission factors and monetary penalty factors than SCR system over the three different layouts. Regarding the emission factors meeting the limits of emission regulations, layout-1 presented the best performance; however, layout-2 was the best from the viewpoints of dropping monetary penalty factors.

ACKNOWLEDGMENT
The research reported in this article was jointly supported by the EU-funded project MODALES (grant agreement no. 815189).